arXiv subjects

Alessandro Giusti

Publications and source records attributed to Alessandro Giusti.

47 records · Page 3Linked to original sources

arXiv2021

Training Lightweight CNNs for Human-Nanodrone Proximity Interaction from Small Datasets using Background Randomization

We consider the task of visually estimating the pose of a human from images acquired by a nearby nano-drone; in this context, we propose a data augmentation approach based on synthetic background substitution to learn a lightweight CNN model from a small real-world training set. Experimental results on data from two different labs proves that the approach improves generalization to unseen environments.

cs.CV↗

arXiv2021

Uncertainty-Aware Self-Supervised Learning of Spatial Perception Tasks

We propose a general self-supervised learning approach for spatial perception tasks, such as estimating the pose of an object relative to the robot, from onboard sensor readings. The model is learned from training episodes, by relying on: a continuous state estimate, possibly inaccurate and affected by odometry drift; and a detector, that sporadically provides supervision about the target pose. We demonstrate the general approach in three different concrete scenarios: a simulated robot arm that visually estimates the pose of an object of interest; a small differential drive robot using 7 infrared sensors to localize a nearby wall; an omnidirectional mobile robot that localizes itself in an environment from camera images. Quantitative results show that the approach works well in all three scenarios, and that explicitly accounting for uncertainty yields statistically significant performance improvements.

cs.RO↗

arXiv2021

Fully Onboard AI-powered Human-Drone Pose Estimation on Ultra-low Power Autonomous Flying Nano-UAVs

Artificial intelligence-powered pocket-sized air robots have the potential to revolutionize the Internet-of-Things ecosystem, acting as autonomous, unobtrusive, and ubiquitous smart sensors. With a few cm$^{2}$ form-factor, nano-sized unmanned aerial vehicles (UAVs) are the natural befit for indoor human-drone interaction missions, as the pose estimation task we address in this work. However, this scenario is challenged by the nano-UAVs' limited payload and computational power that severely relegates the onboard brain to the sub-100 mW microcontroller unit-class. Our work stands at the intersection of the novel parallel ultra-low-power (PULP) architectural paradigm and our general development methodology for deep neural network (DNN) visual pipelines, i.e., covering from perception to control. Addressing the DNN model design, from training and dataset augmentation to 8-bit quantization and deployment, we demonstrate how a PULP-based processor, aboard a nano-UAV, is sufficient for the real-time execution (up to 135 frame/s) of our novel DNN, called PULP-Frontnet. We showcase how, scaling our model's memory and computational requirement, we can significantly improve the onboard inference (top energy efficiency of 0.43 mJ/frame) with no compromise in the quality-of-result vs. a resource-unconstrained baseline (i.e., full-precision DNN). Field experiments demonstrate a closed-loop top-notch autonomous navigation capability, with a heavily resource-constrained 27-gram Crazyflie 2.1 nano-quadrotor. Compared against the control performance achieved using an ideal sensing setup, onboard relative pose inference yields excellent drone behavior in terms of median absolute errors, such as positional (onboard: 41 cm, ideal: 26 cm) and angular (onboard: 3.7$^{\circ}$, ideal: 4.1$^{\circ}$).

cs.RO↗

arXiv2020

Semantic Segmentation on Swiss3DCities: A Benchmark Study on Aerial Photogrammetric 3D Pointcloud Dataset

We introduce a new outdoor urban 3D pointcloud dataset, covering a total area of 2.7 $km^2$, sampled from three Swiss cities with different characteristics. The dataset is manually annotated for semantic segmentation with per-point labels, and is built using photogrammetry from images acquired by multirotors equipped with high-resolution cameras. In contrast to datasets acquired with ground LiDAR sensors, the resulting point clouds are uniformly dense and complete, and are useful to disparate applications, including autonomous driving, gaming and smart city planning. As a benchmark, we report quantitative results of PointNet++, an established point-based deep 3D semantic segmentation model; on this model, we additionally study the impact of using different cities for model generalization.

cs.CV↗

arXiv2020

Learning to predict metal deformations in hot-rolling processes

Hot-rolling is a metal forming process that produces a workpiece with a desired target cross-section from an input workpiece through a sequence of plastic deformations; each deformation is generated by a stand composed of opposing rolls with a specific geometry. In current practice, the rolling sequence (i.e., the sequence of stands and the geometry of their rolls) needed to achieve a given final cross-section is designed by experts based on previous experience, and iteratively refined in a costly trial-and-error process. Finite Element Method simulations are increasingly adopted to make this process more efficient and to test potential rolling sequences, achieving good accuracy at the cost of long simulation times, limiting the practical use of the approach. We propose a supervised learning approach to predict the deformation of a given workpiece by a set of rolls with a given geometry; the model is trained on a large dataset of procedurally-generated FEM simulations, which we publish as supplementary material. The resulting predictor is four orders of magnitude faster than simulations, and yields an average Jaccard Similarity Index of 0.972 (against ground truth from simulations) and 0.925 (against real-world measured deformations); we additionally report preliminary results on using the predictor for automatic planning of rolling sequences.

cs.CE↗

arXiv2019

Vision-based Control of a Quadrotor in User Proximity: Mediated vs End-to-End Learning Approaches

We consider the task of controlling a quadrotor to hover in front of a freely moving user, using input data from an onboard camera. On this specific task we compare two widespread learning paradigms: a mediated approach, which learns an high-level state from the input and then uses it for deriving control signals; and an end-to-end approach, which skips high-level state estimation altogether. We show that despite their fundamental difference, both approaches yield equivalent performance on this task. We finally qualitatively analyze the behavior of a quadrotor implementing such approaches.

cs.RO↗

arXiv2019

Learning Ground Traversability from Simulations

Mobile ground robots operating on unstructured terrain must predict which areas of the environment they are able to pass in order to plan feasible paths. We address traversability estimation as a heightmap classification problem: we build a convolutional neural network that, given an image representing the heightmap of a terrain patch, predicts whether the robot will be able to traverse such patch from left to right. The classifier is trained for a specific robot model (wheeled, tracked, legged, snake-like) using simulation data on procedurally generated training terrains; the trained classifier can be applied to unseen large heightmaps to yield oriented traversability maps, and then plan traversable paths. We extensively evaluate the approach in simulation on six real-world elevation datasets, and run a real-robot validation in one indoor and one outdoor environment.

cs.RO↗

arXiv2019

Learning Long-Range Perception Using Self-Supervision from Short-Range Sensors and Odometry

We introduce a general self-supervised approach to predict the future outputs of a short-range sensor (such as a proximity sensor) given the current outputs of a long-range sensor (such as a camera); we assume that the former is directly related to some piece of information to be perceived (such as the presence of an obstacle in a given position), whereas the latter is information-rich but hard to interpret directly. We instantiate and implement the approach on a small mobile robot to detect obstacles at various distances using the video stream of the robot's forward-pointing camera, by training a convolutional neural network on automatically-acquired datasets. We quantitatively evaluate the quality of the predictions on unseen scenarios, qualitatively evaluate robustness to different operating conditions, and demonstrate usage as the sole input of an obstacle-avoidance controller. We additionally instantiate the approach on a different simulated scenario with complementary characteristics, to exemplify the generality of our contribution.

cs.RO↗

arXiv2014

Assessment of algorithms for mitosis detection in breast cancer histopathology images

The proliferative activity of breast tumors, which is routinely estimated by counting of mitotic figures in hematoxylin and eosin stained histology sections, is considered to be one of the most important prognostic markers. However, mitosis counting is laborious, subjective and may suffer from low inter-observer agreement. With the wider acceptance of whole slide images in pathology labs, automatic image analysis has been proposed as a potential solution for these issues. In this paper, the results from the Assessment of Mitosis Detection Algorithms 2013 (AMIDA13) challenge are described. The challenge was based on a data set consisting of 12 training and 11 testing subjects, with more than one thousand annotated mitotic figures by multiple observers. Short descriptions and results from the evaluation of eleven methods are presented. The top performing method has an error rate that is comparable to the inter-observer agreement among pathologists.

cs.CV↗

arXiv2013

A Fast Learning Algorithm for Image Segmentation with Max-Pooling Convolutional Networks

We present a fast algorithm for training MaxPooling Convolutional Networks to segment images. This type of network yields record-breaking performance in a variety of tasks, but is normally trained on a computationally expensive patch-by-patch basis. Our new method processes each training image in a single pass, which is vastly more efficient. We validate the approach in different scenarios and report a 1500-fold speed-up. In an application to automated steel defect detection and segmentation, we obtain excellent performance with short training times.

cs.CV↗

arXiv2013

Fast Image Scanning with Deep Max-Pooling Convolutional Neural Networks

Deep Neural Networks now excel at image classification, detection and segmentation. When used to scan images by means of a sliding window, however, their high computational complexity can bring even the most powerful hardware to its knees. We show how dynamic programming can speedup the process by orders of magnitude, even when max-pooling layers are present.

cs.CV↗